Forums → Games → Gemcraft CS: 1024 hardcore upgrading

Are you still working on this?

Was very confused with the diagram by cronos51. Did my own version in Illustrator that is a little more colorful and "hopefully" easier to understand.

I'm also working on an "workflow" for the 1024-combine method that one can memorize for those of us who do not like scripting. Did something similar for the 64-combine method in another thread.

Hope this gets more people interested in supergemming. So far I got a pure mana G91e by 1024-combining, but that was "only" 23% better than the G91e that you by 64-combining... Also a question from my side, for the very high grades of triple gems R/O/B, when the blood-bound component has a higher impact, do amplifiers with pure gems still make a difference? I get around 1e23 mana per hit with the triple gem, while one amplifier gem of same grade adds only 1e14 mana/hit. It might be more cost effective to add towers with R/O/B-gems shooting at the mana-farm of permafrozen monsters... Alas, the lag :-(

3.885 billion hits on my displayed Grade 75 manafarm gem. Its got a displayed 1.29125649211714e22 mana per hit. Its got six pure orange Grade 76 amplifier gems. All upgrades were done using the 64 gem upgrade, though I did a 1:63 ROB:OB ratio through Grade 35 on the trap's gem in order to "squeeze" the red out, and make the Orange and Black a little more powerful.

I pull one amplifier gem out, and the displayed mana leech becomes 1.10914824485199e22.
I pull two amplifier gems out, and the displayed mana leech becomes 9.2703999758833e21.
I pull a third amplifier gem out, and the displayed mana leech becomes 7.44931750321675e21.
A fourth amplifier gem removed : 5.62823503056517e21.
A fifth amplifier gem removed : 3.80715255791359e21.
All six amplifier gems removed : 1.98607008526201e21. This is the base amount of displayed mana leech for the gem in the trap, without any amplifiers.

Those six amplifiers brought my displayed mana leech from just under 2e21 to 12.9e21. That's more than a six-fold increase.

Are you upgrading your amplifier gems with the 64 gem upgrade (or 1024 gem upgrade) the way you are the main manafarm gem? If not, you should be.

As a side note, I do not believe that the specials display bug has been fixed. Chain hit (which should not get that high), mana leech, critical damage multiplier, etc. all go to scientific notation two "e" levels early, but they are bugged to start it at e21. So, though the base gem-in-trap without amps shows just under 2e21, it is actually 2e19. The full value, with all six amps, is not 1.29e22, it is really 1.29e20. This display bug is annoying, but does not impact what I've said above, because all seven values are suffering from the same misplaced decimal point bug. That bug, however, is why I specified "displayed mana leech" rather than just calling it mana leech.

I did all my upgrades -- trap gem as well as amplifier gems -- with the 64-combine method so that the amplifiers are 1 grade higher (or four grades lower) than the amplified gem. I was just guessing, that with a blood-bound multiplier of ~10k, the mana-leech value of the hitting gem is a factor of 10k higher then the pure amplifier gem (e.g. 1e22 vs 1e18). So if you add
6x 1e18 to 1e22, it still is 1e22 and the effort and cost of maintaining the high grade amplifier gems would not be worth it.

But apparently, the order of multiplying is different than I thought. If you gain so much by amplifiers, then they first add to the mana-leech value of the hitting gem and after that the blood-bound multiplier is applied. How cool! In this case it is worth to try how 1024-combined trap+amplifier gems perform close to G100e, especially for the killing gem.

I would still love to break endurance without scripting and more than 12 hours of real time effort in at least one level. Started a run on F1 yesterday ;-)

I believe, though I could be wrong, that amplifiers are added before the Bloodbound/Poolbound multipliers are applied.

Nice job for the diagram, MrMait.
In the beginning my goal was the same as yours - a diagram anybody could understand and use, too. Then it took me incredibly long to translate psoreks numbers-only-scheme - and even longer to find the mistakes I had made. More than once I was close to the point of surrender.
I stopped when I thought there shouldn't remain any mistakes and I could build a script. I agree, its quite confusing

Amplifiers:
You and Astro are right, mana-leech is applied before the bloodboundmultiplier. By the way: the 'bonus from amplifiers' that is displayed is, confusingly enough, the same for towers and traps. If the gem is in a trap it has to be multiplied by 4.8 (if trap-skill is maxed; then multiplied again by the bloodboundmultiplier) to get the 'real' amp-bonus.

1024 managem:
If I recall it right for the managem: I think that both amplifiers and managem showed about 1.5 more mana at G100. I don't remember the combined advantage.
I think using 1024 doesn't makes any sense if one is concerned about realtime. Even if the process of upgrading would take the same time per G (which it doesn't), the advantage still would be eaten by the fact of being able to upgrade less often (as oLaudix somewhere proved in some diagrams, too). It might make sense if one would split the process of upgrading in like 20+ steps, which would be a logistical masterpiece to do.

For the killgem I remember the following:
1024 didn't seem to make toooo much sense. While the specials increase, maximum damage heavily decreases.

Don't know where else to complain about this, but I'm frustrated with the bug-riddled mismatch between displayed values of mana-leech and the actual mana gained per frame. Did some tests with my current F1-run:

I compared two different gems in a trap without amplifiers and ~13k frozen monsters on top:
a) triple R/O/B gem starting from a 32-spec and 8x 1024-combined to G69, then "u" upgraded 4x to a final G73, with a displayed mana-leech of 2.332e21

b) dual O/B gem starting from a 32-spec (red resubstituted with orange) 7x 1024-combined to a G69, with a displayed mana-leech of 2.145e15

I used the pause-button to advance 1 frame and got the following values:

number of hits
a) -- 13829
b) -- 2
mana gained by advancing one frame
a) -- 1.156e24
b) -- 2.004e16
mana gained per hit
a) -- 8.384e19
b) -- 1.002e16
ratio between expected mana gained per hit from displayed values
a) -- 4.67%
b) -- 467%

This means without chain-hitting, one actually gains almost 5 times more mana than displayed, and with chain-hitting one gains 1/20 times what is expected from the displayed value ?!? A possible explanation for the latter is, that the mana-leech per hit from chain-hitting decreases with each consecutive hit. That might explain why the expected value is so much higher than the actual gain. But why is mana gain actually higher than displayed for non-chainhitting gems...

I'm really confused right now :-(

Er, to specify, you have less chain then # of monsters, equal chain vs # monsters, or more than # monsters?

Less than = with equal chain would be higher, obviously.

More than = you need to count exactly how many you hit(how much the hit counter) goes up by, and multiply expected gain by that and not the chain value displayed.

Weird bug if you did do it correctly.

@thunderrider:
13829 for a) and 2 for b) are the numbers the respective gem-hit counter went up. Which is reasonable, because a) has a chain-hit length around 260000 and both gems have more than 30 shots-per-second. So the shot-cap is 2 per frame for non-chain-hitting gems and 1 per frame for chain-hitting gems. I had expected the hit counter for the triple gem to go up by 2x13829...

I did an experiment, too. Completely strange all that:
2950 monsters
56.300 chainhitlength

I get 3,5999 times the mana per frame expected, so probably it was 3.6.
That doesn't seem to make any sense. Even less, if compared to the results of TheMaIT ...

Another try seems to suggest:

Traps (with R in the gem) are limited to 1 shot per frame.
Result is multiplied by the 'Mana gain mana pool multiplier'.
Can anybody confirm that?

I got these results with 997 monsters and 1897 chainhitlength.
The results of TheMaiT could suggest:
His 'Mana ... multiplier was at 4.xx. The result for OB could be correct while for ROB with extremely high chainhitlengths something goes wrong? Or there was a little mistake in the calculation ...

Yeah, I got it:
It's all about the displaybug. Once in E+xx mana, one has to divide the figure shown by 100 to get the actual result. That's why TheMaITs results looked so strange
I had been thinking that this multiplier only affects the mana gained by killing monsters, though. Sort of funny that nobody among us ever remarked that. Or did I just miss something thats clear to all of you?

By the way, maybe the mana shard thing is not an exploit but only affected by this multiplier, too. And talismans don't affect that too much or even at all?

Wait. That can't be right. Peter STATED on his developer blog that mana pool multiplier no longer works on multiplying orange.

He said something like "By the way, the mana pool will no longer multiply the mana stolen by orange gems. You'll have to find another way to get huge mana farms :P"

Uh. That's all I gotta say for the 1024 diagrams. Makes zero sense to me. Any help deciphering it into easy-to-follow parentheses or am I gonna have to do it on my own?